Development of a new metric to identify rare patterns in association analysis: The case of analyzing diabetes complications

作者：

Highlights：

• This research aims to address the rare item detection problem in association rule mining.

• A new assessment metric, called adjusted_support, is proposed for rare items detection.

• A large size dataset with the data of about 600,000 patients is used to test the proposed metric.

摘要

•This research aims to address the rare item detection problem in association rule mining.•A new assessment metric, called adjusted_support, is proposed for rare items detection.•A large size dataset with the data of about 600,000 patients is used to test the proposed metric.•Adjusted_support is applied to discover rare association rules for diabetes complications.•Comorbidity index of diabetic patients in various demographic groups is analyzed.

论文关键词：Data mining,Diabetes,Comorbidity,Association rule mining,Rare-pattern identification,Adjusted_support

论文评审过程：Received 19 June 2017, Revised 10 September 2017, Accepted 28 September 2017, Available online 29 September 2017, Version of Record 5 November 2017.

论文官网地址：https://doi.org/10.1016/j.eswa.2017.09.061